Prediction of premixed, n-heptane and iso-octane unopposed jet flames using a reduced kinetic model based on constituents and light species

Abstract

A model of steady, quasi one-dimensional premixed laminar jet flame developing unopposed into a uniform flow has been formulated using a previously successful reduced chemical-kinetics model [10] and [11]. A detailed derivation of the steady quasi one-dimensional conservation equations revealed that it is only under very restrictive conditions – probably very difficult to achieve experimentally and the validity of which is not reported in detail in experimental studies – that the quasi one-dimensional concept is meaningful. The governing equations have been mathematically manipulated to be consistent with the framework of the reduced chemical-kinetics model which relied on constituents representing the heavy species, and on quasi-steady light species and unsteady light species. The flame model includes accurate transport property calculation for high-pressure conditions and a real-gas equation of state. Based on a found self-similarity [10] and [11] which deteriorates at increasingly rich conditions, the chemistry model consists of tables of kinetic rates, quasi-steady species molar fractions and the heavy species mean molar mass extracted from the LLNL model in the framework of the reduced kinetics. The progress variables are only the mass fractions of the unsteady light species and the temperature. The values of the dependent variables are specified at the inflow location and null gradients are specified at the outflow. Simulations were performed for both n-heptane and iso-octane air oxidation over a wide range of pressures and equivalence ratios. The limited documentation of experimental conditions not specifying the inflow velocity (or flux) made it impossible to use this data for detailed comparison. In the one case where the inflow velocity was available for a burner experiment, those conditions were adopted for the simulation and the configuration was changed to a constant-area jet to approach the burner configuration. Results from this simulation compared favorably with the data, considering the different configurations. Results from parametric studies not associated with experimental data showed that at stoichiometric conditions the flame temperature, flame velocity and strain rate are not sensitive to the pressure, although flames become increasingly thinner with increasing pressure and the yield of the unsteady light species is different. Computations conducted at 40 bar for various equivalence ratios and for velocities differing with the equivalence ratio showed that the maximum flame velocity, flame strain and flame temperature were obtained at stoichometric conditions. Finally, we discuss the limitations of utilizing a priori obtained reduced chemical-kinetic models in flames calculations.

Additional details